Author |
Message |
hububalli
Joined: Feb 15, 2014 Posts: 41 Location: UK
|
Posted: Sat Aug 03, 2019 8:45 am Post subject:
VC Spring Reverb issues - overload! |
|
|
I have been messing around with this for months now and have discovered more of whats going on but need some help to get any further.
Feeding a 10 Vpp 550 hz saw in to the input. At pin 3 of U1, the waveform has been attenuated down to 900 mV. At the output pin 5, there is a constant 6 V DC when the input pot is zero. When input pot is full there is a 10.6 Vpp distorted saw wave that is offset by 6v. After C3 220uf cap, the offset is -300 mV.
You can hear the distortion coming on almost straight away from the reverb tank itself as you start turning up the input pot! The overload led comes on almost straight away.
Pin 3, U1.
Pin 5, U1 (5 V Div)
After C3, output to reverb tank (2 V Div)
Secondly, when you get to mixing the original and reverb signals at U5-B, the gain is massive!
With the reverb level pot at zero and original level at max. At the junction between C12 and R20, there is a clean 5 Vpp saw, offset by -400 mV. With the value of R21 at 270k, the gain is huge so the output is massively distorted, maxing out at 24 Vpp.
On MFOS it states;
Quote: | The output of both VCAs go to the inputs (via C10 and C12) of output mixer U5-B and associated components. This is a simple inverting mixer that applies a gain of one to the reverb level VCA's channel and a gain of 10 for the original level VCA's gain. U5-B's output is cap coupled (via non-polarized 10uf C9) to J1, the output jack.
|
Which doesn't seem to match what I am seeing.
Any help would be most appreciated,
http://musicfromouterspace.com/analogsynth_new/VC_REVERB/VC_REVERB.php
Main schematic
Overload Indicator
|
|
Back to top
|
|
|
LukeLabs
Joined: Jan 11, 2020 Posts: 3 Location: Sydney, Australia
|
Posted: Sat Jan 11, 2020 11:26 pm Post subject:
|
|
|
This is a long shot, as it was 6 years ago, but did you ever get a resolution to this?
I've just built the VC Reverb also and I'm having the exact same problem. |
|
Back to top
|
|
|
hububalli
Joined: Feb 15, 2014 Posts: 41 Location: UK
|
Posted: Tue Jan 14, 2020 2:39 pm Post subject:
|
|
|
Hi, Nope never did. tried changing so many components but never got it to work properly. I build the music thing modular spring reverb instead in the end.
Sorry couldn't help further |
|
Back to top
|
|
|
Psynth
Joined: Jul 18, 2018 Posts: 34 Location: UK
Audio files: 8
|
Posted: Wed Jan 15, 2020 2:28 am Post subject:
|
|
|
I've not built this, but the late Ray Wilson's designs are normally well verified and documented. Looking at the schematic though, I think he may have forgotten to say that the circuit is intended for use with an electric guitar or line level instrument as input, not a 10V PP synth signal.
If you follow the signal paths, from input to the reverb, you have U5 a unity buffer, R9/R39 a divide by 11 voltage divider, and then U1, which has a default gain of 20. Taken together, that is a gain of just under 2 on that path. If the input is 10VPP, that would mean U1 would be trying to provide a 20VPP output, but its supply voltage is only +12V (with the other rail being at 0V). Result - a distorted (clipped) signal.
On the 'Original Level' path, with the VCA section at max it is unity gain, then U5-B has a gain of 27. So again, a 10VPP signal is trying to create 270V at U5-B's ouput (pin 7).
Retracing those two paths with a 200mV singal (guitars vary but that is typical), you'd get a 400mVPP signal out of U1, and a 5.4VPP signal out of U5-B, both within the supply voltage parameters.
I think there is a small error on the schemativc around U5-B - according to the text description, R21 should be 100k, not 270k. I suspect a mix between whether the input was to be for guitar or line level during the design.
So, the answer to your problems should be to reduce the overall gains. I'd suggest replacing R24 with a 10k resistor, making the overall gain of the reverb pathway 0.2. That should mean a 1VPP signal output at U1 pin 5.
Note that the LM386 ouput is automatically biased to half the supply voltage, so your 6VDC on it is correct. (Hence the massive 220u blocvking capacitor.) That means you've only got 6V headroom for the output signal. If changing R24 gives you an undistorted signal there, you could push it up a bit, to get say a 4VPP signal, depending on how strong the signal is at C10/R18 - I think you are looking for between 5 and 10VPP there.
Also, in addition to reducing R24 to 100k, increase R20 to 100k, which should prevent the clipping of the original signal.
Peter |
|
Back to top
|
|
|
LukeLabs
Joined: Jan 11, 2020 Posts: 3 Location: Sydney, Australia
|
|
Back to top
|
|
|
artilect99
Joined: Oct 01, 2018 Posts: 49 Location: USA
|
|
Back to top
|
|
|
|